------------------------------------------------------------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  C:\Users\mgolder\Dropbox\campaign_sentiment_paper\replication\data_analysis\sentiment2.log
  log type:  text
 opened on:  14 Jan 2019, 15:27:38

. #delimit ;
delimiter now ;
. *     ***************************************************************** *;
. *     ***************************************************************** *;
. *       File-Name:      sentiment2.do                                   *;
. *       Date:           April 22, 2017                                  *;
. *       Author:         MRG/CC                                          *;
. *       Purpose:        Data analysis replication file for the Online   *;
. *                       Appendix that accompanies, It's Not Only What   *;
. *                       You Say, It's Also How You Say It: The          *;
. *                       Strategic Use Of Campaign Sentiment. Journal of *;
. *                       Politics. Forthcoming                           *;
. *           Input File:     sentiment.dta                                   *;
. *       Output File:    sentiment2.log                                  *;
. *       Data Output:    none                                            *;
.              *       Previous file:  sentiment.dta                                   *;
. *       Machine:        desktop/laptop                                                  *;
. *     ****************************************************************  *;
. *     ****************************************************************  *;
. *     ****************************************************************  *;
. *        Load data                                                      *;
. *     ****************************************************************  *;
. use "C:\Users\mgolder\Dropbox\campaign_sentiment_paper\replication\data_analysis\sentiment.dta", clear;

. set more off;

. desc;

Contains data from C:\Users\mgolder\Dropbox\campaign_sentiment_paper\replication\data_analysis\sentiment.dta
  obs:           421                          
 vars:            25                          22 Apr 2017 15:16
 size:       153,665                          
------------------------------------------------------------------------------------------------------------------------------------------------
              storage   display    value
variable name   type    format     label      variable label
------------------------------------------------------------------------------------------------------------------------------------------------
cow             float   %8.0g                 cow country code
country         str34   %34s                  country name
year            int     %8.0g                 year
election_id     long    %8.0g      election_id
                                              election-specific id
party_name_en~h str120  %120s                 party name in English
family_name     str19   %19s                  party family name (parlgov)
family_id       byte    %8.0g                 party family id (parlgov)
language        long    %10.0g     language   language
positive_sent~t float   %8.0g                 positive emotion minus negative emotion (liwc)
positive_emot~n float   %8.0g                 percent positive words in manifesto (liwc)
negative_emot~n float   %8.0g                 percent negative words in manifesto (liwc)
incumbent       byte    %8.0g                 1 if in government, 0 otherwise (Glasgow Golder 2015
pm              byte    %8.0g                 1 if prime ministerial party, 0 otherwise (Glasgow Golder 2015)
fm              byte    %8.0g                 1 if finance ministerial party, 0 otherwise (parlgov)
left_right      float   %8.0g                 0-10 left-right scale, expert surveys (parlgov)
inflation       float   %8.0g                 inflation rate (WDI 2012, World Bank)
inflation_l     float   %8.0g                 inflation rate lagged one year (WDI 2012, World Bank)
unemployment    float   %8.0g                 unemployment rate (IMF 2015)
unemployment_l  float   %8.0g                 unemployment rate lagged one year (IMF 2015)
rgdpe_growth    float   %9.0g                 growth in real gdp in expenditures (millions of 2011 US dollars)
lrgdpe_growth   float   %9.0g                  lagged growth in real gdp expenditures (millions of 2011 US dollars)
rgdpo_growth    float   %9.0g                 growth in real gdp in output (millions of 2011 US dollars)
lrgdpo_growth   float   %9.0g                 lagged growth in real gdp output (millions of 2011 US dollars)
party_id        int     %8.0g                 party id (parlgov)
party_name      str124  %124s                 party name in native language
------------------------------------------------------------------------------------------------------------------------------------------------
Sorted by: country  year  election_id

. sum;

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
         cow |        421    234.2874    38.36343        200        325
     country |          0
        year |        421    1996.347    9.457847       1980       2012
 election_id |        421    33.21615    21.18554          1         70
party_name~h |          0
-------------+---------------------------------------------------------
 family_name |          0
   family_id |        412    14.24515    9.051924          3         40
    language |        421    3.491686    2.110755          1          7
positive_s~t |        421    1.696912    1.446166       -.68        7.6
positive_e~n |        421     3.02133    1.908511        .64       9.62
-------------+---------------------------------------------------------
negative_e~n |        421    1.324418    .7893864          0       5.22
   incumbent |        421    .2541568    .4359042          0          1
          pm |        421    .1543943    .3617562          0          1
          fm |        421    .1496437    .3571463          0          1
  left_right |        382     4.98458    2.366081         .5   9.824561
-------------+---------------------------------------------------------
   inflation |        400    4.367223    4.699949    -.83553   25.10564
 inflation_l |        400    4.911179    5.451015  -.9461664   28.78333
unemployment |        421    8.899962    4.497289      3.359      22.64
unemployme~l |        414    8.747821    4.522593      3.137       22.9
rgdpe_growth |        421    2.852763    3.571232  -4.305958   12.30451
-------------+---------------------------------------------------------
lrgdpe_gro~h |        421    2.716662    3.391096  -6.225039   10.43808
rgdpo_growth |        421    2.955977    3.621164  -6.712398   12.85193
lrgdpo_gro~h |        421    2.572339    3.704223  -7.322986   10.34733
    party_id |        421    797.2565    465.0217         11       1647
  party_name |          0

. *     ****************************************************************  *;
. *       Generate additional variables that we will need for our         *;
. *       analysis.                                                       *;
. *     ****************************************************************  *;
. gen incumbent_inflation_l = incumbent*inflation_l;
(21 missing values generated)

. label var incumbent_inflation "incumbent*inflation_l";

. gen incumbent_unemployment_l = incumbent*unemployment_l;
(7 missing values generated)

. label var incumbent_unemployment_l "incumbent*unemployment_l";

. gen incumbent_lrgpde_growth = incumbent*lrgdpe_growth;

. label var incumbent_lrgpde_growth "incumbent*lrgpde_growth";

. gen incumbent_lrgpdo_growth = incumbent*lrgdpo_growth;

. label var incumbent_lrgpdo_growth "incumbent*lrgpdo_growth";

. gen incumbent_pm = incumbent*pm;

. label var incumbent_pm "incumbent*pm";

. gen incumbent_fm = incumbent*fm;

. label var incumbent_fm "incumbent*fm";

. gen incumbent_pm_fm = incumbent_pm*incumbent_fm;

. label var incumbent_pm_fm "incumbent_pm*incumbent_fm";

. gen left_right2=left_right*left_right;
(39 missing values generated)

. label var left_right2 "left_right*left_right";

. gen extremist = 0;

. replace extremist=1 if family_id ==14 | family_id==40;
(87 real changes made)

. replace extremist = . if family_id==.;
(9 real changes made, 9 to missing)

. label var extremist "1 if communist/socialist or far right, 0 otherwise";

. gen extremist2 = 0;

. replace extremist2=1 if family_id ==14 | family_id==40 | family_id==19;
(124 real changes made)

. replace extremist2 = . if family_id==.;
(9 real changes made, 9 to missing)

. label var extremist2 "1 if communist/socialist, far right, or greens, 0 otherwise";

. desc;

Contains data from C:\Users\mgolder\Dropbox\campaign_sentiment_paper\replication\data_analysis\sentiment.dta
  obs:           421                          
 vars:            35                          22 Apr 2017 15:16
 size:       170,505                          
------------------------------------------------------------------------------------------------------------------------------------------------
              storage   display    value
variable name   type    format     label      variable label
------------------------------------------------------------------------------------------------------------------------------------------------
cow             float   %8.0g                 cow country code
country         str34   %34s                  country name
year            int     %8.0g                 year
election_id     long    %8.0g      election_id
                                              election-specific id
party_name_en~h str120  %120s                 party name in English
family_name     str19   %19s                  party family name (parlgov)
family_id       byte    %8.0g                 party family id (parlgov)
language        long    %10.0g     language   language
positive_sent~t float   %8.0g                 positive emotion minus negative emotion (liwc)
positive_emot~n float   %8.0g                 percent positive words in manifesto (liwc)
negative_emot~n float   %8.0g                 percent negative words in manifesto (liwc)
incumbent       byte    %8.0g                 1 if in government, 0 otherwise (Glasgow Golder 2015
pm              byte    %8.0g                 1 if prime ministerial party, 0 otherwise (Glasgow Golder 2015)
fm              byte    %8.0g                 1 if finance ministerial party, 0 otherwise (parlgov)
left_right      float   %8.0g                 0-10 left-right scale, expert surveys (parlgov)
inflation       float   %8.0g                 inflation rate (WDI 2012, World Bank)
inflation_l     float   %8.0g                 inflation rate lagged one year (WDI 2012, World Bank)
unemployment    float   %8.0g                 unemployment rate (IMF 2015)
unemployment_l  float   %8.0g                 unemployment rate lagged one year (IMF 2015)
rgdpe_growth    float   %9.0g                 growth in real gdp in expenditures (millions of 2011 US dollars)
lrgdpe_growth   float   %9.0g                  lagged growth in real gdp expenditures (millions of 2011 US dollars)
rgdpo_growth    float   %9.0g                 growth in real gdp in output (millions of 2011 US dollars)
lrgdpo_growth   float   %9.0g                 lagged growth in real gdp output (millions of 2011 US dollars)
party_id        int     %8.0g                 party id (parlgov)
party_name      str124  %124s                 party name in native language
incumbent_inf~l float   %9.0g                 incumbent*inflation_l
incumbent_une~l float   %9.0g                 incumbent*unemployment_l
incumb~e_growth float   %9.0g                 incumbent*lrgpde_growth
incumb~o_growth float   %9.0g                 incumbent*lrgpdo_growth
incumbent_pm    float   %9.0g                 incumbent*pm
incumbent_fm    float   %9.0g                 incumbent*fm
incumbent_pm_fm float   %9.0g                 incumbent_pm*incumbent_fm
left_right2     float   %9.0g                 left_right*left_right
extremist       float   %9.0g                 1 if communist/socialist or far right, 0 otherwise
extremist2      float   %9.0g                 1 if communist/socialist, far right, or greens, 0 otherwise
------------------------------------------------------------------------------------------------------------------------------------------------
Sorted by: country  year  election_id
     Note: Dataset has changed since last saved.

. sum;

    Variable |        Obs        Mean    Std. Dev.       Min        Max
-------------+---------------------------------------------------------
         cow |        421    234.2874    38.36343        200        325
     country |          0
        year |        421    1996.347    9.457847       1980       2012
 election_id |        421    33.21615    21.18554          1         70
party_name~h |          0
-------------+---------------------------------------------------------
 family_name |          0
   family_id |        412    14.24515    9.051924          3         40
    language |        421    3.491686    2.110755          1          7
positive_s~t |        421    1.696912    1.446166       -.68        7.6
positive_e~n |        421     3.02133    1.908511        .64       9.62
-------------+---------------------------------------------------------
negative_e~n |        421    1.324418    .7893864          0       5.22
   incumbent |        421    .2541568    .4359042          0          1
          pm |        421    .1543943    .3617562          0          1
          fm |        421    .1496437    .3571463          0          1
  left_right |        382     4.98458    2.366081         .5   9.824561
-------------+---------------------------------------------------------
   inflation |        400    4.367223    4.699949    -.83553   25.10564
 inflation_l |        400    4.911179    5.451015  -.9461664   28.78333
unemployment |        421    8.899962    4.497289      3.359      22.64
unemployme~l |        414    8.747821    4.522593      3.137       22.9
rgdpe_growth |        421    2.852763    3.571232  -4.305958   12.30451
-------------+---------------------------------------------------------
lrgdpe_gro~h |        421    2.716662    3.391096  -6.225039   10.43808
rgdpo_growth |        421    2.955977    3.621164  -6.712398   12.85193
lrgdpo_gro~h |        421    2.572339    3.704223  -7.322986   10.34733
    party_id |        421    797.2565    465.0217         11       1647
  party_name |          0
-------------+---------------------------------------------------------
incumben~n_l |        400    1.359408    3.844508  -.9461664   28.78333
incumben~t_l |        414    2.160932    4.305807          0       22.9
inc~e_growth |        421    .6976881     2.05097  -6.225039   10.43808
inc~o_growth |        421    .6570862    2.107362  -7.322986   10.34733
incumbent_pm |        421    .1543943    .3617562          0          1
-------------+---------------------------------------------------------
incumbent_fm |        421    .1377672    .3450656          0          1
incumbe~m_fm |        421    .1045131    .3062889          0          1
 left_right2 |        382    30.42972    24.22462        .25     96.522
   extremist |        412     .211165    .4086314          0          1
  extremist2 |        412    .3009709    .4592377          0          1

. xtset language;
       panel variable:  language (unbalanced)

. *     ****************************************************************  *;
. *       Online Appendix A: Table 1 -- Check finance minister            *;
. *     ****************************************************************  *;
. tab pm fm if incumbent==1;

1 if prime |
ministeria |
l party, 0 |
 otherwise |     1 if finance
  (Glasgow | ministerial party, 0
    Golder |  otherwise (parlgov)
     2015) |         0          1 |     Total
-----------+----------------------+----------
         0 |        28         14 |        42 
         1 |        21         44 |        65 
-----------+----------------------+----------
     Total |        49         58 |       107 


. * Model 1*;
. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_fm, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       421
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1015                         Obs per group: min =        36
       between = 0.0044                                        avg =      60.1
       overall = 0.0286                                        max =       102

                                                Wald chi2(2)       =     57.88
corr(u_i, Xb)  = 0.0109                         Prob > chi2        =    0.0000

                            (Replications based on 70 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .4847589   .0824133     5.88   0.000     .3232318    .6462861
incumbent_fm |   .0794286   .0866416     0.92   0.359    -.0903857     .249243
       _cons |   1.562765   .1502123    10.40   0.000     1.268354    1.857175
-------------+----------------------------------------------------------------
     sigma_u |  1.3461966
     sigma_e |  .68890528
         rho |  .79246849   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. * Model 2 *;
. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm incumbent_fm, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       421
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1097                         Obs per group: min =        36
       between = 0.0783                                        avg =      60.1
       overall = 0.0389                                        max =       102

                                                Wald chi2(3)       =     61.73
corr(u_i, Xb)  = 0.0366                         Prob > chi2        =    0.0000

                            (Replications based on 70 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .3604294   .0837544     4.30   0.000     .1962738    .5245849
incumbent_pm |   .2839577   .0793587     3.58   0.000     .1284174     .439498
incumbent_fm |  -.0116512   .0914236    -0.13   0.899    -.1908382    .1675358
       _cons |    1.56307   .1502828    10.40   0.000     1.268521    1.857619
-------------+----------------------------------------------------------------
     sigma_u |  1.3401241
     sigma_e |  .68659681
         rho |   .7920852   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. * Model 3 *;
. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm incumbent_fm incumbent_pm_fm, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       421
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1101                         Obs per group: min =        36
       between = 0.0881                                        avg =      60.1
       overall = 0.0413                                        max =       102

                                                Wald chi2(4)       =     79.06
corr(u_i, Xb)  = 0.0431                         Prob > chi2        =    0.0000

                               (Replications based on 70 clusters in election_id)
---------------------------------------------------------------------------------
                |   Observed   Bootstrap                         Normal-based
positive_sent~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------+----------------------------------------------------------------
      incumbent |   .3851388   .0996004     3.87   0.000     .1899255     .580352
   incumbent_pm |   .2253779    .109063     2.07   0.039     .0116184    .4391374
   incumbent_fm |  -.0856149   .1351416    -0.63   0.526    -.3504875    .1792578
incumbent_pm_fm |   .1249361   .1914192     0.65   0.514    -.2502386    .5001109
          _cons |   1.562967   .1502989    10.40   0.000     1.268386    1.857547
----------------+----------------------------------------------------------------
        sigma_u |   1.339067
        sigma_e |  .68728823
            rho |  .79149314   (fraction of variance due to u_i)
---------------------------------------------------------------------------------

. *     ****************************************************************  *;
. *       No evidence that finance minister matters, PM party still       *;
. *       matters.                                                        *;
. *     ****************************************************************  *;
. *     ****************************************************************  *;
. *       Online Appendix B: Replication can be found in                  *;
. *       figure1_appendix.do                                             *;
. *     ****************************************************************  *;
. *     ****************************************************************  *;
. *       Online Appendix C: Fixed effects                                *;
. *     ****************************************************************  *;
. *     ****************************************************************  *;
. *       Least squares dummy variable version of the fixed effects model *;
. *       Table 2: Language fixed effects                                 *;
. *     ****************************************************************  *;
. tabulate language, gen(l);

   language |      Freq.     Percent        Cum.
------------+-----------------------------------
      Dutch |        102       24.23       24.23
    English |         85       20.19       44.42
     French |         36        8.55       52.97
     German |         46       10.93       63.90
    Italian |         54       12.83       76.72
 Portuguese |         50       11.88       88.60
    Spanish |         48       11.40      100.00
------------+-----------------------------------
      Total |        421      100.00

. bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent l1 l2 l3 l4 l5 l6 l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        421
                                                Replications      =        400
                                                Wald chi2(8)      =    4321.69
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9064
                                                Adj R-squared     =     0.9046
                                                Root MSE          =     0.6884

                            (Replications based on 70 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .5280646   .0695955     7.59   0.000     .3916599    .6644694
          l1 |   .2724351    .058642     4.65   0.000      .157499    .3873713
          l2 |   2.236972   .1091718    20.49   0.000     2.022999    2.450944
          l3 |   1.414651   .1156081    12.24   0.000     1.188063    1.641238
          l4 |   1.578583   .1020915    15.46   0.000     1.378488    1.778679
          l5 |   .3119116   .1011545     3.08   0.002     .1136523    .5101708
          l6 |   4.256064   .1985606    21.43   0.000     3.866893    4.645236
          l7 |   1.807863   .0395302    45.73   0.000     1.730385    1.885341
------------------------------------------------------------------------------

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 2918.36
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm l1 l2 l3 l4 l5 l6 l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        421
                                                Replications      =        400
                                                Wald chi2(9)      =    4528.54
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9073
                                                Adj R-squared     =     0.9053
                                                Root MSE          =     0.6858

                            (Replications based on 70 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |    .356483   .0750253     4.75   0.000     .2094362    .5035298
incumbent_pm |   .2800309   .0749149     3.74   0.000     .1332004    .4268614
          l1 |   .2842901   .0577556     4.92   0.000     .1710912    .3974889
          l2 |   2.231964   .1070136    20.86   0.000     2.022221    2.441707
          l3 |   1.403095   .1159216    12.10   0.000     1.175893    1.630298
          l4 |   1.590935    .101993    15.60   0.000     1.391033    1.790838
          l5 |    .324112   .0964818     3.36   0.001     .1350112    .5132128
          l6 |   4.246839   .1970182    21.56   0.000      3.86069    4.632987
          l7 |   1.787529   .0398179    44.89   0.000     1.709487     1.86557
------------------------------------------------------------------------------

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 2768.42
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm left_right left_right2  l1 l2 l3 l4 l5 l
> 6 l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        382
                                                Replications      =        400
                                                Wald chi2(11)     =    6529.70
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9265
                                                Adj R-squared     =     0.9243
                                                Root MSE          =     0.6279

                            (Replications based on 69 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .1545325   .0637387     2.42   0.015     .0296071     .279458
incumbent_pm |    .239634   .0733366     3.27   0.001     .0958969    .3833711
  left_right |   .4104397   .0690804     5.94   0.000     .2750446    .5458347
 left_right2 |  -.0376656   .0069866    -5.39   0.000    -.0513591   -.0239722
          l1 |  -.4676069   .1525842    -3.06   0.002    -.7666665   -.1685474
          l2 |   1.347853   .1704191     7.91   0.000     1.013838    1.681869
          l3 |   .6298614   .1741129     3.62   0.000     .2886064    .9711164
          l4 |   .7813827   .1953624     4.00   0.000     .3984793    1.164286
          l5 |  -.5159477   .1702469    -3.03   0.002    -.8496254   -.1822699
          l6 |   3.628111   .1922918    18.87   0.000     3.251226    4.004996
          l7 |   .8839384   .1542622     5.73   0.000       .58159    1.186287
------------------------------------------------------------------------------

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 1115.74
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist  l1 l2 l3 l4 l5 l6 l7, noconst
> ant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        412
                                                Replications      =        400
                                                Wald chi2(10)     =    4430.47
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9150
                                                Adj R-squared     =     0.9129
                                                Root MSE          =     0.6628

                            (Replications based on 70 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2555482   .0776734     3.29   0.001     .1033112    .4077852
incumbent_pm |   .2428077   .0770869     3.15   0.002     .0917202    .3938951
   extremist |  -.5040679   .1044126    -4.83   0.000    -.7087128    -.299423
          l1 |   .4067363   .0596712     6.82   0.000      .289783    .5236897
          l2 |   2.326127   .0968879    24.01   0.000      2.13623    2.516023
          l3 |    1.57546   .1257885    12.52   0.000     1.328919    1.822001
          l4 |   1.747294   .1046427    16.70   0.000     1.542198     1.95239
          l5 |   .4653904    .125027     3.72   0.000     .2203419    .7104389
          l6 |   4.459227   .1700462    26.22   0.000     4.125943    4.792512
          l7 |   1.897445   .0496511    38.22   0.000      1.80013    1.994759
------------------------------------------------------------------------------

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 2691.76
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist inflation_l l1 l2 l3 l4 l5 l6 
> l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
......................................x...........   250
..................................................   300
.......................................x..........   350
..................................................   400

Linear regression                               Number of obs     =        391
                                                Replications      =        398
                                                Wald chi2(11)     =    3834.27
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9171
                                                Adj R-squared     =     0.9147
                                                Root MSE          =     0.6550

                            (Replications based on 64 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2650156   .0774989     3.42   0.001     .1131204    .4169107
incumbent_pm |   .2062344   .0806363     2.56   0.011     .0481901    .3642786
   extremist |  -.4949045   .1078627    -4.59   0.000    -.7063115   -.2834975
 inflation_l |  -.0281764   .0131099    -2.15   0.032    -.0538714   -.0024814
          l1 |   .4924331   .0648037     7.60   0.000     .3654203     .619446
          l2 |   2.411103    .110231    21.87   0.000     2.195055    2.627152
          l3 |   1.688937   .1196473    14.12   0.000     1.454433    1.923442
          l4 |   1.821056    .147628    12.34   0.000      1.53171    2.110401
          l5 |   .6522199   .1281546     5.09   0.000     .4010415    .9033982
          l6 |   4.793165   .1283879    37.33   0.000     4.541529      5.0448
          l7 |   2.039242   .0927275    21.99   0.000       1.8575    2.220985
------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 2050.39
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist unemployment_l l1 l2 l3 l4 l5 
> l6 l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        405
                                                Replications      =        400
                                                Wald chi2(11)     =    4051.87
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9214
                                                Adj R-squared     =     0.9192
                                                Root MSE          =     0.6372

                              (Replications based on 68 clusters in election_id)
--------------------------------------------------------------------------------
               |   Observed   Bootstrap                         Normal-based
positive_sen~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
     incumbent |   .2443843    .073679     3.32   0.001     .0999761    .3887925
  incumbent_pm |   .2395275   .0797955     3.00   0.003     .0831313    .3959238
     extremist |  -.4540733    .112523    -4.04   0.000    -.6746144   -.2335322
unemployment_l |   -.027279   .0109923    -2.48   0.013    -.0488236   -.0057344
            l1 |   .5366903   .0786618     6.82   0.000      .382516    .6908646
            l2 |   2.598977    .113233    22.95   0.000     2.377044    2.820909
            l3 |   1.801089   .1583418    11.37   0.000     1.490745    2.111433
            l4 |   1.974813    .133668    14.77   0.000     1.712829    2.236798
            l5 |   .7039958   .1730659     4.07   0.000     .3647928    1.043199
            l6 |   4.732812   .1687713    28.04   0.000     4.402026    5.063597
            l7 |   2.342591   .2005628    11.68   0.000     1.949495    2.735687
--------------------------------------------------------------------------------

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 1275.05
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist lrgdpe_growth l1 l2 l3 l4 l5 l
> 6 l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        412
                                                Replications      =        400
                                                Wald chi2(11)     =    4050.71
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9150
                                                Adj R-squared     =     0.9127
                                                Root MSE          =     0.6636

                             (Replications based on 70 clusters in election_id)
-------------------------------------------------------------------------------
              |   Observed   Bootstrap                         Normal-based
positive_se~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    incumbent |   .2557025   .0779149     3.28   0.001     .1029922    .4084129
 incumbent_pm |   .2426094    .077017     3.15   0.002     .0916589      .39356
    extremist |  -.5041256   .1043136    -4.83   0.000    -.7085764   -.2996748
lrgdpe_growth |  -.0014959   .0105937    -0.14   0.888    -.0222592    .0192675
           l1 |   .4098158   .0638657     6.42   0.000     .2846414    .5349902
           l2 |   2.331603   .1071012    21.77   0.000     2.121688    2.541517
           l3 |   1.578512   .1293168    12.21   0.000     1.325056    1.831968
           l4 |   1.750964   .1081822    16.19   0.000     1.538931    1.962997
           l5 |   .4695301   .1315262     3.57   0.000     .2117436    .7273166
           l6 |   4.464276   .1801493    24.78   0.000      4.11119    4.817362
           l7 |   1.901297   .0656753    28.95   0.000     1.772576    2.030019
-------------------------------------------------------------------------------

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 1621.22
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist inflation_l incumbent_inflatio
> n_l l1 l2 l3 l4 l5 l6 l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
......................................x...........   250
..................................................   300
.......................................x..........   350
..................................................   400

Linear regression                               Number of obs     =        391
                                                Replications      =        398
                                                Wald chi2(12)     =    3862.98
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9189
                                                Adj R-squared     =     0.9164
                                                Root MSE          =     0.6488

                                     (Replications based on 64 clusters in election_id)
---------------------------------------------------------------------------------------
                      |   Observed   Bootstrap                         Normal-based
   positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------------+----------------------------------------------------------------
            incumbent |   .0522727   .1128402     0.46   0.643      -.16889    .2734355
         incumbent_pm |   .2331414   .0805509     2.89   0.004     .0752646    .3910182
            extremist |  -.4713886   .1034543    -4.56   0.000    -.6741554   -.2686218
          inflation_l |  -.0404941   .0144799    -2.80   0.005    -.0688742   -.0121139
incumbent_inflation_l |   .0379998   .0158918     2.39   0.017     .0068525    .0691472
                   l1 |   .5461171   .0700022     7.80   0.000     .4089154    .6833188
                   l2 |   2.454976   .1123859    21.84   0.000     2.234703    2.675248
                   l3 |   1.730096   .1211941    14.28   0.000      1.49256    1.967632
                   l4 |    1.87685   .1518218    12.36   0.000     1.579285    2.174416
                   l5 |   .6970959   .1297811     5.37   0.000     .4427296    .9514623
                   l6 |   4.864135   .1335382    36.43   0.000     4.602405    5.125865
                   l7 |   2.092006   .1013188    20.65   0.000     1.893425    2.290587
---------------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 1897.37
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist unemployment_l incumbent_unemp
> loyment_l l1 l2 l3 l4 l5 l6 l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        405
                                                Replications      =        400
                                                Wald chi2(12)     =    4112.47
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9216
                                                Adj R-squared     =     0.9192
                                                Root MSE          =     0.6376

                                        (Replications based on 68 clusters in election_id)
------------------------------------------------------------------------------------------
                         |   Observed   Bootstrap                         Normal-based
      positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
               incumbent |   .3445268   .1381043     2.49   0.013     .0738472    .6152063
            incumbent_pm |   .2563192    .080232     3.19   0.001     .0990673     .413571
               extremist |  -.4538443   .1124822    -4.03   0.000    -.6743054   -.2333832
          unemployment_l |  -.0240029   .0121882    -1.97   0.049    -.0478914   -.0001144
incumbent_unemployment_l |  -.0128492   .0166773    -0.77   0.441    -.0455361    .0198378
                      l1 |   .5091963   .0873468     5.83   0.000     .3379998    .6803928
                      l2 |   2.572505   .1174265    21.91   0.000     2.342353    2.802657
                      l3 |   1.770857   .1659549    10.67   0.000     1.445591    2.096123
                      l4 |   1.945749   .1401316    13.89   0.000     1.671096    2.220401
                      l5 |    .673814   .1750612     3.85   0.000     .3307004    1.016928
                      l6 |   4.704872   .1756963    26.78   0.000     4.360514    5.049231
                      l7 |   2.306851   .2107452    10.95   0.000     1.893798    2.719904
------------------------------------------------------------------------------------------

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 1208.06
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist lrgdpe_growth incumbent_lrgpde
> _growth l1 l2 l3 l4 l5 l6 l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        412
                                                Replications      =        400
                                                Wald chi2(12)     =    4076.33
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9150
                                                Adj R-squared     =     0.9125
                                                Root MSE          =     0.6643

                                       (Replications based on 70 clusters in election_id)
-----------------------------------------------------------------------------------------
                        |   Observed   Bootstrap                         Normal-based
     positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------+----------------------------------------------------------------
              incumbent |   .2306242   .0957496     2.41   0.016     .0429585    .4182899
           incumbent_pm |   .2422159   .0782696     3.09   0.002     .0888103    .3956214
              extremist |  -.5055037   .1031806    -4.90   0.000    -.7077339   -.3032735
          lrgdpe_growth |  -.0037399   .0113967    -0.33   0.743    -.0260769    .0185972
incumbent_lrgpde_growth |   .0091339    .018501     0.49   0.622    -.0271273    .0453951
                     l1 |   .4160387   .0640868     6.49   0.000     .2904308    .5416466
                     l2 |   2.337025    .106009    22.05   0.000     2.129251    2.544799
                     l3 |   1.584946   .1289622    12.29   0.000     1.332184    1.837707
                     l4 |   1.757179   .1090193    16.12   0.000     1.543506    1.970853
                     l5 |   .4768854   .1341342     3.56   0.000     .2139873    .7397836
                     l6 |   4.472407   .1809022    24.72   0.000     4.117845    4.826969
                     l7 |   1.907839   .0676405    28.21   0.000     1.775266    2.040412
-----------------------------------------------------------------------------------------

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 1600.17
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist inflation_l incumbent_inflatio
> n_l unemployment_l incumbent_unemployment_l lrgdpe_growth incumbent_lrgpde_growth l1 l2 l3 l4 l5 l6 l7, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
...................................x..............   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        388
                                                Replications      =        399
                                                Wald chi2(16)     =    4070.31
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9240
                                                Adj R-squared     =     0.9207
                                                Root MSE          =     0.6301

                                        (Replications based on 63 clusters in election_id)
------------------------------------------------------------------------------------------
                         |   Observed   Bootstrap                         Normal-based
      positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
               incumbent |   .1365095   .1570403     0.87   0.385    -.1712838    .4443027
            incumbent_pm |   .2422422   .0848937     2.85   0.004     .0758535    .4086309
               extremist |  -.4310096   .1001717    -4.30   0.000    -.6273426   -.2346766
             inflation_l |  -.0354643   .0143035    -2.48   0.013    -.0634987     -.00743
   incumbent_inflation_l |   .0314988   .0175639     1.79   0.073    -.0029258    .0659234
          unemployment_l |  -.0256988   .0128596    -2.00   0.046    -.0509032   -.0004944
incumbent_unemployment_l |    -.01094   .0162829    -0.67   0.502    -.0428539    .0209739
           lrgdpe_growth |  -.0183259   .0112593    -1.63   0.104    -.0403938     .003742
 incumbent_lrgpde_growth |   .0128794    .018253     0.71   0.480    -.0228959    .0486547
                      l1 |   .6787212   .0946771     7.17   0.000     .4931574     .864285
                      l2 |    2.76053   .1376733    20.05   0.000     2.490696    3.030365
                      l3 |   1.959835   .1599573    12.25   0.000     1.646325    2.273346
                      l4 |    2.10871   .1792676    11.76   0.000     1.757352    2.460068
                      l5 |   .9444995   .1911418     4.94   0.000     .5698684     1.31913
                      l6 |   5.106086    .173769    29.38   0.000     4.765505    5.446667
                      l7 |   2.549758   .2294983    11.11   0.000      2.09995    2.999567
------------------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

. testparm l1 l2 l3 l4 l5 l6 l7;

 ( 1)  l1 = 0
 ( 2)  l2 = 0
 ( 3)  l3 = 0
 ( 4)  l4 = 0
 ( 5)  l5 = 0
 ( 6)  l6 = 0
 ( 7)  l7 = 0

           chi2(  7) = 1434.85
         Prob > chi2 =    0.0000

.  *     ****************************************************************  *;
. *       Least squares dummy variable version of the fixed effects model *;
. *       Table 3: Country fixed effects                                  *;
. *     ****************************************************************  *;
. tabulate cow, gen(c);

cow country |
       code |      Freq.     Percent        Cum.
------------+-----------------------------------
        200 |         40        9.50        9.50
        205 |         45       10.69       20.19
        210 |        102       24.23       44.42
        220 |         36        8.55       52.97
        230 |         48       11.40       64.37
        235 |         50       11.88       76.25
        255 |         46       10.93       87.17
        325 |         54       12.83      100.00
------------+-----------------------------------
      Total |        421      100.00

. bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent c1 c2 c3 c4 c5 c6 c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        421
                                                Replications      =        400
                                                Wald chi2(9)      =    5064.62
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9091
                                                Adj R-squared     =     0.9071
                                                Root MSE          =     0.6792

                            (Replications based on 70 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .5500852   .0686345     8.01   0.000      .415564    .6846063
          c1 |   2.504735   .0971137    25.79   0.000     2.314396    2.695075
          c2 |   1.988194   .1187203    16.75   0.000     1.755506    2.220881
          c3 |    .267038   .0584944     4.57   0.000     .1523911    .3816848
          c4 |   1.409145   .1165679    12.09   0.000     1.180676    1.637614
          c5 |   1.803734   .0392937    45.90   0.000      1.72672    1.880748
          c6 |    4.25078     .19842    21.42   0.000     3.861884    4.639676
          c7 |   1.569967   .1021638    15.37   0.000     1.369729    1.770204
          c8 |   .3070181   .1018952     3.01   0.003     .1073071    .5067291
------------------------------------------------------------------------------

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 3339.91
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm c1 c2 c3 c4 c5 c6 c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        421
                                                Replications      =        400
                                                Wald chi2(10)     =    5207.29
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9098
                                                Adj R-squared     =     0.9076
                                                Root MSE          =     0.6775

                            (Replications based on 70 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .4019289   .0718474     5.59   0.000     .2611105    .5427473
incumbent_pm |   .2402137   .0734176     3.27   0.001     .0963179    .3841095
          c1 |   2.488625   .0965828    25.77   0.000     2.299326    2.677924
          c2 |   1.994875   .1181536    16.88   0.000     1.763298    2.226451
          c3 |   .2774454   .0576059     4.82   0.000       .16454    .3903508
          c4 |   1.399476   .1167752    11.98   0.000     1.170601    1.628351
          c5 |   1.786473   .0397513    44.94   0.000     1.708562    1.864384
          c6 |   4.243099   .1971743    21.52   0.000     3.856644    4.629553
          c7 |   1.580943   .1020011    15.50   0.000     1.381024    1.780861
          c8 |   .3176997   .0976132     3.25   0.001     .1263814     .509018
------------------------------------------------------------------------------

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 3186.40
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm left_right left_right2  c1 c2 c3 c4 c5 c
> 6 c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        382
                                                Replications      =        400
                                                Wald chi2(12)     =    9457.75
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9285
                                                Adj R-squared     =     0.9262
                                                Root MSE          =     0.6199

                            (Replications based on 69 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2020487   .0654425     3.09   0.002     .0737838    .3303135
incumbent_pm |    .200573   .0751872     2.67   0.008     .0532088    .3479373
  left_right |   .3925604   .0678528     5.79   0.000     .2595714    .5255494
 left_right2 |  -.0361531   .0068549    -5.27   0.000    -.0495884   -.0227178
          c1 |   1.655928   .1513364    10.94   0.000     1.359314    1.952542
          c2 |   1.185612   .1789647     6.62   0.000     .8348472    1.536376
          c3 |  -.4336103   .1508137    -2.88   0.004    -.7291998   -.1380209
          c4 |    .665514   .1734594     3.84   0.000     .3255398    1.005488
          c5 |   .9277124   .1522272     6.09   0.000     .6293526    1.226072
          c6 |   3.665507   .1909906    19.19   0.000     3.291172    4.039842
          c7 |   .8130647   .1933887     4.20   0.000     .4340297      1.1921
          c8 |  -.4810966   .1693183    -2.84   0.004    -.8129544   -.1492387
------------------------------------------------------------------------------

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 1463.70
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist  c1 c2 c3 c4 c5 c6 c7 c8, noco
> nstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        412
                                                Replications      =        400
                                                Wald chi2(11)     =    4982.70
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9166
                                                Adj R-squared     =     0.9143
                                                Root MSE          =     0.6574

                            (Replications based on 70 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2963222   .0779488     3.80   0.000     .1435453    .4490991
incumbent_pm |   .2118489   .0769733     2.75   0.006     .0609839    .3627139
   extremist |  -.4830776    .105601    -4.57   0.000    -.6900518   -.2761033
          c1 |   2.532873   .0906985    27.93   0.000     2.355107    2.710638
          c2 |   2.131735   .1234397    17.27   0.000     1.889798    2.373672
          c3 |     .39617   .0599525     6.61   0.000     .2786653    .5136747
          c4 |    1.56517   .1262337    12.40   0.000     1.317756    1.812583
          c5 |   1.892106   .0494555    38.26   0.000     1.795175    1.989037
          c6 |   4.447458   .1706736    26.06   0.000     4.112944    4.781972
          c7 |   1.732833   .1050076    16.50   0.000     1.527022    1.938645
          c8 |   .4538519   .1256209     3.61   0.000     .2076394    .7000644
------------------------------------------------------------------------------

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 2941.15
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist inflation_l c1 c2 c3 c4 c5 c6 
> c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................x...............................    50
..................x...............................   100
..................................................   150
.................................................x   200
......................................x...........   250
..................................................   300
.......................................x..........   350
..................................................   400

Linear regression                               Number of obs     =        391
                                                Replications      =        395
                                                Wald chi2(12)     =    4394.39
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9183
                                                Adj R-squared     =     0.9157
                                                Root MSE          =     0.6513

                            (Replications based on 64 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2980686   .0788276     3.78   0.000     .1435692    .4525679
incumbent_pm |    .184273   .0804317     2.29   0.022     .0266298    .3419162
   extremist |  -.4792741   .1079142    -4.44   0.000     -.690782   -.2677661
 inflation_l |  -.0269638   .0133514    -2.02   0.043    -.0531321   -.0007955
          c1 |   2.602358    .122112    21.31   0.000     2.363023    2.841693
          c2 |    2.25105    .142837    15.76   0.000     1.971095    2.531005
          c3 |   .4800842   .0661769     7.25   0.000     .3503799    .6097885
          c4 |   1.675578   .1204285    13.91   0.000     1.439542    1.911613
          c5 |   2.028608   .0931609    21.78   0.000     1.846016      2.2112
          c6 |   4.769229   .1340808    35.57   0.000     4.506436    5.032023
          c7 |   1.805943   .1489736    12.12   0.000      1.51396    2.097926
          c8 |   .6347887    .130631     4.86   0.000     .3787567    .8908208
------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 5 bootstrap replicates;
      standard-error estimates include only complete replications.

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 2158.29
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist unemployment_l c1 c2 c3 c4 c5 
> c6 c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................x...............   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        405
                                                Replications      =        399
                                                Wald chi2(12)     =    5035.34
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9223
                                                Adj R-squared     =     0.9199
                                                Root MSE          =     0.6346

                              (Replications based on 68 clusters in election_id)
--------------------------------------------------------------------------------
               |   Observed   Bootstrap                         Normal-based
positive_sen~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
     incumbent |   .2762496   .0737833     3.74   0.000      .131637    .4208622
  incumbent_pm |   .2149281   .0789164     2.72   0.006     .0602548    .3696015
     extremist |  -.4401334   .1130436    -3.89   0.000    -.6616947   -.2185721
unemployment_l |  -.0189575   .0101903    -1.86   0.063    -.0389302    .0010152
            c1 |   2.676222   .1180787    22.66   0.000     2.444792    2.907652
            c2 |   2.363497   .1339556    17.64   0.000     2.100949    2.626045
            c3 |   .4874944   .0777201     6.27   0.000     .3351657     .639823
            c4 |   1.721722   .1518964    11.33   0.000     1.424011    2.019434
            c5 |   2.201397   .1793176    12.28   0.000     1.849941    2.552853
            c6 |   4.668074   .1651181    28.27   0.000     4.344449      4.9917
            c7 |    1.89542   .1276319    14.85   0.000     1.645266    2.145574
            c8 |   .6202983   .1743387     3.56   0.000     .2786008    .9619958
--------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 1478.71
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist lrgdpe_growth c1 c2 c3 c4 c5 c
> 6 c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        412
                                                Replications      =        400
                                                Wald chi2(12)     =    5088.53
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9166
                                                Adj R-squared     =     0.9141
                                                Root MSE          =     0.6581

                             (Replications based on 70 clusters in election_id)
-------------------------------------------------------------------------------
              |   Observed   Bootstrap                         Normal-based
positive_se~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    incumbent |   .2971161   .0787264     3.77   0.000     .1428152    .4514169
 incumbent_pm |    .211462   .0771387     2.74   0.006      .060273     .362651
    extremist |  -.4823036   .1058157    -4.56   0.000    -.6896986   -.2749086
lrgdpe_growth |   .0039833   .0097296     0.41   0.682    -.0150863     .023053
           c1 |   2.524399   .1022236    24.69   0.000     2.324045    2.724754
           c2 |   2.111409   .1331539    15.86   0.000     1.850432    2.372385
           c3 |   .3876575   .0650479     5.96   0.000      .260166    .5151489
           c4 |   1.556738   .1294447    12.03   0.000     1.303031    1.810445
           c5 |   1.881688    .059085    31.85   0.000     1.765884    1.997493
           c6 |   4.433666   .1791184    24.75   0.000       4.0826    4.784731
           c7 |   1.722634   .1087405    15.84   0.000     1.509506    1.935761
           c8 |   .4424874   .1298938     3.41   0.001     .1879002    .6970746
-------------------------------------------------------------------------------

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 2001.31
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist inflation_l incumbent_inflatio
> n_l c1 c2 c3 c4 c5 c6 c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................x...............................    50
..................x...............................   100
..................................................   150
.................................................x   200
......................................x...........   250
..................................................   300
.......................................x..........   350
..................................................   400

Linear regression                               Number of obs     =        391
                                                Replications      =        395
                                                Wald chi2(13)     =    4394.51
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9200
                                                Adj R-squared     =     0.9173
                                                Root MSE          =     0.6452

                                     (Replications based on 64 clusters in election_id)
---------------------------------------------------------------------------------------
                      |   Observed   Bootstrap                         Normal-based
   positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------------+----------------------------------------------------------------
            incumbent |   .0881791   .1122757     0.79   0.432    -.1318773    .3082354
         incumbent_pm |   .2113076   .0803322     2.63   0.009     .0538594    .3687558
            extremist |  -.4566032   .1038464    -4.40   0.000    -.6601385    -.253068
          inflation_l |  -.0390969   .0147757    -2.65   0.008    -.0680567   -.0101372
incumbent_inflation_l |    .037328   .0156048     2.39   0.017     .0067433    .0679128
                   c1 |   2.640201   .1240883    21.28   0.000     2.396992    2.883409
                   c2 |   2.298543   .1441402    15.95   0.000     2.016033    2.581052
                   c3 |   .5331582   .0712221     7.49   0.000     .3935654     .672751
                   c4 |   1.716376   .1221159    14.06   0.000     1.477033    1.955718
                   c5 |   2.080731    .101523    20.50   0.000     1.881749    2.279712
                   c6 |   4.839603    .138418    34.96   0.000     4.568308    5.110897
                   c7 |   1.861166   .1534171    12.13   0.000     1.560475    2.161858
                   c8 |   .6793502   .1325236     5.13   0.000     .4196088    .9390916
---------------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 5 bootstrap replicates;
      standard-error estimates include only complete replications.

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 2021.80
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist unemployment_l incumbent_unemp
> loyment_l c1 c2 c3 c4 c5 c6 c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................x...............   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        405
                                                Replications      =        399
                                                Wald chi2(13)     =    5023.18
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9224
                                                Adj R-squared     =     0.9198
                                                Root MSE          =     0.6349

                                        (Replications based on 68 clusters in election_id)
------------------------------------------------------------------------------------------
                         |   Observed   Bootstrap                         Normal-based
      positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
               incumbent |   .3730707   .1348585     2.77   0.006     .1087528    .6373885
            incumbent_pm |   .2312905   .0799689     2.89   0.004     .0745543    .3880267
               extremist |  -.4399711   .1129934    -3.89   0.000    -.6614341   -.2185082
          unemployment_l |  -.0158211   .0117101    -1.35   0.177    -.0387724    .0071302
incumbent_unemployment_l |  -.0124404   .0165361    -0.75   0.452    -.0448507    .0199698
                      c1 |   2.650263    .127311    20.82   0.000     2.400738    2.899788
                      c2 |   2.338872   .1350791    17.31   0.000     2.074122    2.603622
                      c3 |   .4610847    .087716     5.26   0.000     .2891646    .6330048
                      c4 |    1.69279   .1614542    10.48   0.000     1.376346    2.009234
                      c5 |   2.167396   .1920367    11.29   0.000     1.791011    2.543781
                      c6 |     4.6413   .1727906    26.86   0.000     4.302636    4.979963
                      c7 |   1.867618   .1360148    13.73   0.000     1.601034    2.134203
                      c8 |   .5914335   .1778548     3.33   0.001     .2428445    .9400225
------------------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 1 bootstrap replicate;
      standard-error estimates include only complete replications.

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 1378.36
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist lrgdpe_growth incumbent_lrgpde
> _growth c1 c2 c3 c4 c5 c6 c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Linear regression                               Number of obs     =        412
                                                Replications      =        400
                                                Wald chi2(13)     =    5119.13
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9166
                                                Adj R-squared     =     0.9139
                                                Root MSE          =     0.6587

                                       (Replications based on 70 clusters in election_id)
-----------------------------------------------------------------------------------------
                        |   Observed   Bootstrap                         Normal-based
     positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------+----------------------------------------------------------------
              incumbent |   .2683157   .0956049     2.81   0.005     .0809335     .455698
           incumbent_pm |   .2108865   .0784153     2.69   0.007     .0571953    .3645777
              extremist |  -.4838103   .1046602    -4.62   0.000    -.6889406   -.2786801
          lrgdpe_growth |   .0014132   .0105377     0.13   0.893    -.0192403    .0220667
incumbent_lrgpde_growth |   .0105481   .0183704     0.57   0.566    -.0254573    .0465535
                     c1 |    2.53141   .1015698    24.92   0.000     2.332337    2.730483
                     c2 |   2.116815   .1325272    15.97   0.000     1.857067    2.376564
                     c3 |   .3947578   .0651743     6.06   0.000     .2670185    .5224971
                     c4 |   1.564084   .1289731    12.13   0.000     1.311301    1.816866
                     c5 |   1.889166   .0606853    31.13   0.000     1.770225    2.008107
                     c6 |   4.442936   .1794465    24.76   0.000     4.091228    4.794645
                     c7 |   1.729702   .1092947    15.83   0.000     1.515488    1.943915
                     c8 |   .4508765   .1319166     3.42   0.001     .1923248    .7094282
-----------------------------------------------------------------------------------------

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 1972.42
         Prob > chi2 =    0.0000

.  bootstrap, reps(400) seed(10101) cluster(election_id): reg positive_sentiment incumbent incumbent_pm extremist inflation_l incumbent_inflatio
> n_l unemployment_l incumbent_unemployment_l lrgdpe_growth incumbent_lrgpde_growth c1 c2 c3 c4 c5 c6 c7 c8, noconstant;
(running regress on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
.......................x..........................   100
...................x...............x.........x....   150
..................................................   200
..................................................   250
..................................................   300
............x.....................................   350
..................................................   400

Linear regression                               Number of obs     =        388
                                                Replications      =        395
                                                Wald chi2(17)     =    4389.61
                                                Prob > chi2       =     0.0000
                                                R-squared         =     0.9243
                                                Adj R-squared     =     0.9208
                                                Root MSE          =     0.6298

                                        (Replications based on 63 clusters in election_id)
------------------------------------------------------------------------------------------
                         |   Observed   Bootstrap                         Normal-based
      positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
               incumbent |   .1466886   .1564901     0.94   0.349    -.1600264    .4534036
            incumbent_pm |   .2314505   .0844645     2.74   0.006      .065903     .396998
               extremist |  -.4243037    .099559    -4.26   0.000    -.6194358   -.2291716
             inflation_l |  -.0339712   .0148055    -2.29   0.022    -.0629894    -.004953
   incumbent_inflation_l |   .0316089   .0174925     1.81   0.071    -.0026757    .0658936
          unemployment_l |  -.0196954   .0132233    -1.49   0.136    -.0456127    .0062218
incumbent_unemployment_l |  -.0107553   .0161193    -0.67   0.505    -.0423485    .0208379
           lrgdpe_growth |  -.0146557   .0113655    -1.29   0.197    -.0369317    .0076203
 incumbent_lrgpde_growth |   .0142371   .0183455     0.78   0.438    -.0217193    .0501936
                      c1 |   2.791546   .1459829    19.12   0.000     2.505425    3.077667
                      c2 |   2.592859   .1827109    14.19   0.000     2.234752    2.950966
                      c3 |    .633107   .1017597     6.22   0.000     .4336616    .8325524
                      c4 |   1.890809   .1666127    11.35   0.000     1.564254    2.217364
                      c5 |   2.431499   .2381919    10.21   0.000     1.964651    2.898346
                      c6 |   5.032163   .1769132    28.44   0.000     4.685419    5.378906
                      c7 |   2.040812   .1796955    11.36   0.000     1.688615    2.393008
                      c8 |    .865793   .2005604     4.32   0.000     .4727018    1.258884
------------------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 5 bootstrap replicates;
      standard-error estimates include only complete replications.

. testparm c1 c2 c3 c4 c5 c6 c7 c8;

 ( 1)  c1 = 0
 ( 2)  c2 = 0
 ( 3)  c3 = 0
 ( 4)  c4 = 0
 ( 5)  c5 = 0
 ( 6)  c6 = 0
 ( 7)  c7 = 0
 ( 8)  c8 = 0

           chi2(  8) = 1488.75
         Prob > chi2 =    0.0000

.  *     ****************************************************************  *;
. *       Robustness Check: Different measure of extremist party that     *;
. *       includes greens                                                 *;
. *     ****************************************************************  *;
. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist2, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       412
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1777                         Obs per group: min =        35
       between = 0.0227                                        avg =      58.9
       overall = 0.0381                                        max =       100

                                                Wald chi2(3)       =     72.24
corr(u_i, Xb)  = -0.0201                        Prob > chi2        =    0.0000

                            (Replications based on 70 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2414633   .0710991     3.40   0.001     .1021117     .380815
incumbent_pm |   .2214776   .0785819     2.82   0.005       .06746    .3754953
  extremist2 |  -.4432264   .0855761    -5.18   0.000    -.6109525   -.2755004
       _cons |   1.753465   .1681734    10.43   0.000     1.423851    2.083079
-------------+----------------------------------------------------------------
     sigma_u |  1.3517896
     sigma_e |  .66510351
         rho |  .80510063   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist2 inflation_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       391
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1961                         Obs per group: min =        31
       between = 0.4322                                        avg =      55.9
       overall = 0.0008                                        max =       100

                                                Wald chi2(4)       =     82.77
corr(u_i, Xb)  = -0.2185                        Prob > chi2        =    0.0000

                            (Replications based on 64 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2686315   .0729616     3.68   0.000     .1256294    .4116336
incumbent_pm |   .1776945   .0791991     2.24   0.025     .0224672    .3329219
  extremist2 |  -.4252975   .0917305    -4.64   0.000     -.605086    -.245509
 inflation_l |  -.0309111   .0132669    -2.33   0.020    -.0569138   -.0049084
       _cons |   1.885983    .199779     9.44   0.000     1.494423    2.277542
-------------+----------------------------------------------------------------
     sigma_u |  1.4215917
     sigma_e |  .65898521
         rho |   .8231247   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist2 unemployment_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       405
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1815                         Obs per group: min =        35
       between = 0.0032                                        avg =      57.9
       overall = 0.0200                                        max =       100

                                                Wald chi2(4)       =     82.87
corr(u_i, Xb)  = -0.0798                        Prob > chi2        =    0.0000

                              (Replications based on 68 clusters in election_id)
--------------------------------------------------------------------------------
               |   Observed   Bootstrap                         Normal-based
positive_sen~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
     incumbent |   .2362144   .0689914     3.42   0.001     .1009937    .3714351
  incumbent_pm |   .2201315   .0794568     2.77   0.006      .064399    .3758641
    extremist2 |   -.391203   .0911305    -4.29   0.000    -.5698154   -.2125906
unemployment_l |  -.0270626   .0106544    -2.54   0.011    -.0479449   -.0061803
         _cons |   1.968926   .2077596     9.48   0.000     1.561725    2.376128
---------------+----------------------------------------------------------------
       sigma_u |  1.3896808
       sigma_e |  .64015646
           rho |  .82494745   (fraction of variance due to u_i)
--------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist2 lrgdpe_growth, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       412
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1777                         Obs per group: min =        35
       between = 0.0243                                        avg =      58.9
       overall = 0.0378                                        max =       100

                                                Wald chi2(4)       =     73.34
corr(u_i, Xb)  = -0.0208                        Prob > chi2        =    0.0000

                             (Replications based on 70 clusters in election_id)
-------------------------------------------------------------------------------
              |   Observed   Bootstrap                         Normal-based
positive_se~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    incumbent |   .2415349   .0712809     3.39   0.001     .1018269    .3812429
 incumbent_pm |   .2214173   .0784285     2.82   0.005     .0677003    .3751343
   extremist2 |  -.4431741    .085499    -5.18   0.000    -.6107491   -.2755991
lrgdpe_growth |  -.0005201   .0106731    -0.05   0.961     -.021439    .0203988
        _cons |    1.75486    .169289    10.37   0.000     1.423059     2.08666
--------------+----------------------------------------------------------------
      sigma_u |  1.3519926
      sigma_e |  .66592998
          rho |   .8047578   (fraction of variance due to u_i)
-------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist2 inflation_l incumbent_inflat
> ion_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       391
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.2155                         Obs per group: min =        31
       between = 0.4626                                        avg =      55.9
       overall = 0.0012                                        max =       100

                                                Wald chi2(5)       =    103.06
corr(u_i, Xb)  = -0.2245                        Prob > chi2        =    0.0000

                                     (Replications based on 64 clusters in election_id)
---------------------------------------------------------------------------------------
                      |   Observed   Bootstrap                         Normal-based
   positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------------+----------------------------------------------------------------
            incumbent |   .0409099    .109279     0.37   0.708    -.1732731    .2550929
         incumbent_pm |   .2069457   .0787823     2.63   0.009     .0525353    .3613561
           extremist2 |  -.4090166   .0895386    -4.57   0.000    -.5845091   -.2335241
          inflation_l |  -.0438748   .0145406    -3.02   0.003    -.0723738   -.0153758
incumbent_inflation_l |   .0404131   .0160934     2.51   0.012     .0088706    .0719556
                _cons |   1.941156   .2063595     9.41   0.000     1.536698    2.345613
----------------------+----------------------------------------------------------------
              sigma_u |  1.4292806
              sigma_e |  .65181545
                  rho |  .82783067   (fraction of variance due to u_i)
---------------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist2 unemployment_l incumbent_une
> mployment_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       405
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1824                         Obs per group: min =        35
       between = 0.0038                                        avg =      57.9
       overall = 0.0201                                        max =       100

                                                Wald chi2(5)       =     89.33
corr(u_i, Xb)  = -0.0794                        Prob > chi2        =    0.0000

                                        (Replications based on 68 clusters in election_id)
------------------------------------------------------------------------------------------
                         |   Observed   Bootstrap                         Normal-based
      positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
               incumbent |   .3196099   .1345407     2.38   0.018      .055915    .5833049
            incumbent_pm |   .2342394   .0800548     2.93   0.003     .0773349     .391144
              extremist2 |  -.3898061   .0911289    -4.28   0.000    -.5684154   -.2111967
          unemployment_l |  -.0243533   .0117446    -2.07   0.038    -.0473722   -.0013343
incumbent_unemployment_l |  -.0106591   .0170422    -0.63   0.532    -.0440612    .0227429
                   _cons |   1.944424   .2060247     9.44   0.000     1.540623    2.348225
-------------------------+----------------------------------------------------------------
                 sigma_u |  1.3898869
                 sigma_e |  .64064291
                     rho |  .82477082   (fraction of variance due to u_i)
------------------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist2 lrgdpe_growth incumbent_lrgp
> de_growth, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       412
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1778                         Obs per group: min =        35
       between = 0.0255                                        avg =      58.9
       overall = 0.0377                                        max =       100

                                                Wald chi2(5)       =     86.44
corr(u_i, Xb)  = -0.0214                        Prob > chi2        =    0.0000

                                       (Replications based on 70 clusters in election_id)
-----------------------------------------------------------------------------------------
                        |   Observed   Bootstrap                         Normal-based
     positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------+----------------------------------------------------------------
              incumbent |   .2241229   .0912159     2.46   0.014     .0453431    .4029027
           incumbent_pm |   .2211562   .0795183     2.78   0.005     .0653033    .3770091
             extremist2 |  -.4436204   .0848539    -5.23   0.000     -.609931   -.2773098
          lrgdpe_growth |  -.0020839   .0114413    -0.18   0.855    -.0245085    .0203406
incumbent_lrgpde_growth |   .0063703   .0186663     0.34   0.733    -.0302149    .0429555
                  _cons |   1.759285   .1666978    10.55   0.000     1.432563    2.086006
------------------------+----------------------------------------------------------------
                sigma_u |  1.3522145
                sigma_e |  .66669524
                    rho |  .80444827   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist2 inflation_l incumbent_inflat
> ion_l unemployment_l incumbent_unemployment_l lrgdpe_growth incumbent_lrgpde_growth, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       388
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.2093                         Obs per group: min =        31
       between = 0.3518                                        avg =      55.4
       overall = 0.0000                                        max =       100

                                                Wald chi2(9)       =    109.62
corr(u_i, Xb)  = -0.2516                        Prob > chi2        =    0.0000

                                        (Replications based on 63 clusters in election_id)
------------------------------------------------------------------------------------------
                         |   Observed   Bootstrap                         Normal-based
      positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
               incumbent |   .1212897   .1570188     0.77   0.440    -.1864616    .4290409
            incumbent_pm |   .2165085   .0837775     2.58   0.010     .0523075    .3807094
              extremist2 |  -.3686215   .0900925    -4.09   0.000    -.5451997   -.1920434
             inflation_l |   -.038588   .0143185    -2.69   0.007    -.0666518   -.0105241
   incumbent_inflation_l |   .0333565   .0177403     1.88   0.060    -.0014139    .0681269
          unemployment_l |  -.0256675   .0124026    -2.07   0.038    -.0499763   -.0013588
incumbent_unemployment_l |  -.0091825   .0165341    -0.56   0.579    -.0415889    .0232238
           lrgdpe_growth |   -.018044   .0110777    -1.63   0.103    -.0397558    .0036678
 incumbent_lrgpde_growth |   .0108986   .0179037     0.61   0.543    -.0241921    .0459892
                   _cons |   2.163726   .2145482    10.09   0.000      1.74322    2.584233
-------------------------+----------------------------------------------------------------
                 sigma_u |  1.4552321
                 sigma_e |  .63333057
                     rho |  .84075499   (fraction of variance due to u_i)
------------------------------------------------------------------------------------------

. *     ****************************************************************  *;
. *       Results are robust                                              *;
. *     ****************************************************************  *;
. *     ****************************************************************  *;
. *       Robustness Check: No clustering                                 *;
. *     ****************************************************************  *;
. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       421
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1008                         Obs per group: min =        36
       between = 0.0007                                        avg =      60.1
       overall = 0.0266                                        max =       102

                                                Wald chi2(1)       =     63.54
corr(u_i, Xb)  = 0.0046                         Prob > chi2        =    0.0000

                                (Replications based on clustering on language)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .5280646   .0662458     7.97   0.000     .3982252     .657904
       _cons |   1.562701   .0728621    21.45   0.000     1.419894    1.705508
-------------+----------------------------------------------------------------
     sigma_u |  1.3474834
     sigma_e |  .68835933
         rho |  .79304294   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent incumbent_pm, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       421
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1097                         Obs per group: min =        36
       between = 0.0803                                        avg =      60.1
       overall = 0.0390                                        max =       102

                                                Wald chi2(2)       =     64.62
corr(u_i, Xb)  = 0.0370                         Prob > chi2        =    0.0000

                                (Replications based on clustering on language)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |    .356483   .0803254     4.44   0.000     .1990481    .5139179
incumbent_pm |   .2800309   .1005174     2.79   0.005     .0830204    .4770415
       _cons |   1.563074    .072883    21.45   0.000     1.420226    1.705922
-------------+----------------------------------------------------------------
     sigma_u |    1.34004
     sigma_e |  .68576863
         rho |  .79246181   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent incumbent_pm left_right left_right2, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       382
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1817                         Obs per group: min =        35
       between = 0.0954                                        avg =      54.6
       overall = 0.0675                                        max =        90

                                                Wald chi2(4)       =     81.17
corr(u_i, Xb)  = 0.0652                         Prob > chi2        =    0.0000

                                (Replications based on clustering on language)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .1545325   .0683213     2.26   0.024     .0206253    .2884397
incumbent_pm |    .239634   .0933819     2.57   0.010     .0566088    .4226593
  left_right |   .4104397   .0687056     5.97   0.000     .2757792    .5451001
 left_right2 |  -.0376656   .0069402    -5.43   0.000    -.0512682    -.024063
       _cons |   .7746918   .1572456     4.93   0.000     .4664961    1.082888
-------------+----------------------------------------------------------------
     sigma_u |  1.3907853
     sigma_e |  .62788657
         rho |  .83069045   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent incumbent_pm extremist, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       412
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1834                         Obs per group: min =        35
       between = 0.1569                                        avg =      58.9
       overall = 0.0325                                        max =       100

                                                Wald chi2(3)       =    100.79
corr(u_i, Xb)  = -0.0408                        Prob > chi2        =    0.0000

                                (Replications based on clustering on language)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2555482   .0893261     2.86   0.004     .0804723    .4306242
incumbent_pm |   .2428077   .0928909     2.61   0.009     .0607448    .4248705
   extremist |  -.5040679   .0917198    -5.50   0.000    -.6838353   -.3243004
       _cons |   1.719485      .0821    20.94   0.000     1.558572    1.880398
-------------+----------------------------------------------------------------
     sigma_u |  1.3615153
     sigma_e |  .66276504
         rho |   .8084339   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent incumbent_pm extremist inflation_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       391
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.2057                         Obs per group: min =        31
       between = 0.4541                                        avg =      55.9
       overall = 0.0012                                        max =       100

                                                Wald chi2(4)       =     86.10
corr(u_i, Xb)  = -0.2179                        Prob > chi2        =    0.0000

                                (Replications based on clustering on language)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2650156   .0999538     2.65   0.008     .0691097    .4609215
incumbent_pm |   .2062344   .1111482     1.86   0.064     -.011612    .4240808
   extremist |  -.4949045   .0934682    -5.29   0.000    -.6780988   -.3117102
 inflation_l |  -.0281764   .0120851    -2.33   0.020    -.0518628   -.0044901
       _cons |   1.847508   .0949638    19.45   0.000     1.661382    2.033633
-------------+----------------------------------------------------------------
     sigma_u |  1.4253002
     sigma_e |  .65502756
         rho |  .82562331   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent incumbent_pm extremist unemployment_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       405
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1890                         Obs per group: min =        35
       between = 0.0197                                        avg =      57.9
       overall = 0.0170                                        max =       100

                                                Wald chi2(4)       =     94.31
corr(u_i, Xb)  = -0.0975                        Prob > chi2        =    0.0000

                                  (Replications based on clustering on language)
--------------------------------------------------------------------------------
               |   Observed   Bootstrap                         Normal-based
positive_sen~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
     incumbent |   .2443843   .0862416     2.83   0.005     .0753539    .4134148
  incumbent_pm |   .2395275   .1022416     2.34   0.019     .0391377    .4399174
     extremist |  -.4540733   .0952873    -4.77   0.000    -.6408329   -.2673136
unemployment_l |   -.027279   .0127712    -2.14   0.033      -.05231   -.0022479
         _cons |   1.944063    .127186    15.29   0.000     1.694783    2.193343
---------------+----------------------------------------------------------------
       sigma_u |   1.398648
       sigma_e |  .63723563
           rho |  .82810323   (fraction of variance due to u_i)
--------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent incumbent_pm extremist lrgdpe_growth, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       412
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1835                         Obs per group: min =        35
       between = 0.1683                                        avg =      58.9
       overall = 0.0319                                        max =       100

                                                Wald chi2(4)       =    102.93
corr(u_i, Xb)  = -0.0428                        Prob > chi2        =    0.0000

                                 (Replications based on clustering on language)
-------------------------------------------------------------------------------
              |   Observed   Bootstrap                         Normal-based
positive_se~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
--------------+----------------------------------------------------------------
    incumbent |   .2557025   .0896735     2.85   0.004     .0799458    .4314593
 incumbent_pm |   .2426094   .0932472     2.60   0.009     .0598483    .4253705
    extremist |  -.5041256    .091762    -5.49   0.000    -.6839758   -.3242754
lrgdpe_growth |  -.0014959   .0111905    -0.13   0.894    -.0234289    .0204371
        _cons |   1.723572   .0867041    19.88   0.000     1.553635    1.893509
--------------+----------------------------------------------------------------
      sigma_u |   1.362111
      sigma_e |  .66357166
          rho |  .80819254   (fraction of variance due to u_i)
-------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent incumbent_pm extremist inflation_l incumbent_inflation_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       391
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.2228                         Obs per group: min =        31
       between = 0.4746                                        avg =      55.9
       overall = 0.0015                                        max =       100

                                                Wald chi2(5)       =    102.62
corr(u_i, Xb)  = -0.2237                        Prob > chi2        =    0.0000

                                         (Replications based on clustering on language)
---------------------------------------------------------------------------------------
                      |   Observed   Bootstrap                         Normal-based
   positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------------+----------------------------------------------------------------
            incumbent |   .0522727   .1141615     0.46   0.647    -.1714796    .2760251
         incumbent_pm |   .2331414   .1037805     2.25   0.025     .0297353    .4365476
            extremist |  -.4713886   .0934428    -5.04   0.000    -.6545332    -.288244
          inflation_l |  -.0404941   .0163181    -2.48   0.013     -.072477   -.0085111
incumbent_inflation_l |   .0379998   .0193842     1.96   0.050     7.49e-06    .0759922
                _cons |   1.899274   .1040277    18.26   0.000     1.695384    2.103165
----------------------+----------------------------------------------------------------
              sigma_u |  1.4322672
              sigma_e |  .64877972
                  rho |  .82974785   (fraction of variance due to u_i)
---------------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent incumbent_pm extremist unemployment_l incumbent_unemployment_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       405
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1902                         Obs per group: min =        35
       between = 0.0223                                        avg =      57.9
       overall = 0.0172                                        max =       100

                                                Wald chi2(5)       =     98.12
corr(u_i, Xb)  = -0.0970                        Prob > chi2        =    0.0000

                                            (Replications based on clustering on language)
------------------------------------------------------------------------------------------
                         |   Observed   Bootstrap                         Normal-based
      positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
               incumbent |   .3445268   .1388105     2.48   0.013     .0724631    .6165904
            incumbent_pm |   .2563192    .102466     2.50   0.012     .0554895    .4571488
               extremist |  -.4538443   .0952802    -4.76   0.000      -.64059   -.2670986
          unemployment_l |  -.0240029   .0134528    -1.78   0.074    -.0503699    .0023641
incumbent_unemployment_l |  -.0128492   .0150542    -0.85   0.393    -.0423549    .0166566
                   _cons |   1.915028   .1314023    14.57   0.000     1.657484    2.172572
-------------------------+----------------------------------------------------------------
                 sigma_u |  1.3988999
                 sigma_e |  .63756734
                     rho |  .82800631   (fraction of variance due to u_i)
------------------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101): xtreg positive_sentiment incumbent incumbent_pm extremist lrgdpe_growth incumbent_lrgpde_growth, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       412
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1838                         Obs per group: min =        35
       between = 0.1689                                        avg =      58.9
       overall = 0.0317                                        max =       100

                                                Wald chi2(5)       =    102.40
corr(u_i, Xb)  = -0.0438                        Prob > chi2        =    0.0000

                                           (Replications based on clustering on language)
-----------------------------------------------------------------------------------------
                        |   Observed   Bootstrap                         Normal-based
     positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------+----------------------------------------------------------------
              incumbent |   .2306242   .1019822     2.26   0.024     .0307428    .4305056
           incumbent_pm |   .2422159   .0935906     2.59   0.010     .0587816    .4256501
              extremist |  -.5055037   .0916036    -5.52   0.000    -.6850435   -.3259639
          lrgdpe_growth |  -.0037399   .0137105    -0.27   0.785    -.0306119    .0231322
incumbent_lrgpde_growth |   .0091339   .0211179     0.43   0.665    -.0322564    .0505241
                  _cons |   1.730052    .087367    19.80   0.000     1.558816    1.901288
------------------------+----------------------------------------------------------------
                sigma_u |  1.3624803
                sigma_e |  .66426329
                    rho |  .80795349   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. *     ****************************************************************  *;
. *       Results are robust                                              *;
. *     ****************************************************************  *;
.  *     ****************************************************************  *;
. *       In our response memo, we also ran a model where we assume that  *;
. *       positive sentiment declines linearly from the center of the     *;
. *       policy space.  This is captured by extremist3. We're expecting  *;
. *       a negative and significant coefficient on extremist3.           *;
. *     ****************************************************************  *;
. gen extremist3 = abs(5-left_right);
(39 missing values generated)

. order left_right extremist3, last;

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist3, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       382
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1635                         Obs per group: min =        35
       between = 0.0325                                        avg =      54.6
       overall = 0.0545                                        max =        90

                                                Wald chi2(3)       =     83.87
corr(u_i, Xb)  = 0.0469                         Prob > chi2        =    0.0000

                            (Replications based on 69 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .2021136   .0679765     2.97   0.003      .068882    .3353452
incumbent_pm |   .2277535   .0770167     2.96   0.003     .0768034    .3787035
  extremist3 |  -.1619933   .0305988    -5.29   0.000    -.2219658   -.1020207
       _cons |   2.002718   .1896248    10.56   0.000      1.63106    2.374376
-------------+----------------------------------------------------------------
     sigma_u |  1.3981678
     sigma_e |  .63399944
         rho |  .82945076   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist3 inflation_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       361
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1659                         Obs per group: min =        30
       between = 0.2219                                        avg =      51.6
       overall = 0.0136                                        max =        90

                                                Wald chi2(4)       =     78.21
corr(u_i, Xb)  = -0.0833                        Prob > chi2        =    0.0000

                            (Replications based on 63 clusters in election_id)
------------------------------------------------------------------------------
             |   Observed   Bootstrap                         Normal-based
positive_s~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
   incumbent |   .1977656   .0717046     2.76   0.006     .0572272    .3383041
incumbent_pm |   .1927448   .0848816     2.27   0.023     .0263799    .3591097
  extremist3 |  -.1631062   .0293604    -5.56   0.000    -.2206516   -.1055608
 inflation_l |  -.0202662   .0120058    -1.69   0.091    -.0437971    .0032646
       _cons |   2.097288   .1997777    10.50   0.000     1.705731    2.488845
-------------+----------------------------------------------------------------
     sigma_u |  1.4367683
     sigma_e |  .63253589
         rho |  .83764776   (fraction of variance due to u_i)
------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist3 unemployment_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       376
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1708                         Obs per group: min =        35
       between = 0.0151                                        avg =      53.7
       overall = 0.0371                                        max =        90

                                                Wald chi2(4)       =     88.40
corr(u_i, Xb)  = -0.0019                        Prob > chi2        =    0.0000

                              (Replications based on 67 clusters in election_id)
--------------------------------------------------------------------------------
               |   Observed   Bootstrap                         Normal-based
positive_sen~t |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
---------------+----------------------------------------------------------------
     incumbent |   .1901927   .0666388     2.85   0.004      .059583    .3208024
  incumbent_pm |   .2231743   .0735307     3.04   0.002     .0790567     .367292
    extremist3 |  -.1556946   .0276459    -5.63   0.000    -.2098796   -.1015096
unemployment_l |  -.0217602   .0108841    -2.00   0.046    -.0430926   -.0004278
         _cons |   2.177773   .2239512     9.72   0.000     1.738837    2.616709
---------------+----------------------------------------------------------------
       sigma_u |  1.4198976
       sigma_e |  .62068198
           rho |  .83957138   (fraction of variance due to u_i)
--------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist3 inflation_l incumbent_inflat
> ion_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       361
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1782                         Obs per group: min =        30
       between = 0.2444                                        avg =      51.6
       overall = 0.0139                                        max =        90

                                                Wald chi2(5)       =     73.14
corr(u_i, Xb)  = -0.0895                        Prob > chi2        =    0.0000

                                     (Replications based on 63 clusters in election_id)
---------------------------------------------------------------------------------------
                      |   Observed   Bootstrap                         Normal-based
   positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------------+----------------------------------------------------------------
            incumbent |   .0280014   .0906927     0.31   0.758    -.1497531    .2057559
         incumbent_pm |   .2139747   .0813929     2.63   0.009     .0544476    .3735018
           extremist3 |  -.1619236   .0306012    -5.29   0.000    -.2219009   -.1019463
          inflation_l |  -.0315014    .013166    -2.39   0.017    -.0573063   -.0056966
incumbent_inflation_l |   .0301817   .0143605     2.10   0.036     .0020356    .0583277
                _cons |   2.145504   .2030097    10.57   0.000     1.747613    2.543396
----------------------+----------------------------------------------------------------
              sigma_u |  1.4402986
              sigma_e |  .62876842
                  rho |  .83992705   (fraction of variance due to u_i)
---------------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist3 unemployment_l incumbent_une
> mployment_l, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       376
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1711                         Obs per group: min =        35
       between = 0.0145                                        avg =      53.7
       overall = 0.0370                                        max =        90

                                                Wald chi2(5)       =     89.10
corr(u_i, Xb)  = -0.0023                        Prob > chi2        =    0.0000

                                        (Replications based on 67 clusters in election_id)
------------------------------------------------------------------------------------------
                         |   Observed   Bootstrap                         Normal-based
      positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
               incumbent |   .2401794   .1339114     1.79   0.073     -.022282    .5026409
            incumbent_pm |    .230922    .074541     3.10   0.002     .0848244    .3770196
              extremist3 |  -.1547477   .0275359    -5.62   0.000     -.208717   -.1007783
          unemployment_l |  -.0200802   .0112575    -1.78   0.074    -.0421445    .0019842
incumbent_unemployment_l |  -.0062504    .015957    -0.39   0.695    -.0375255    .0250247
                   _cons |   2.160678   .2197984     9.83   0.000     1.729881    2.591475
-------------------------+----------------------------------------------------------------
                 sigma_u |  1.4201539
                 sigma_e |  .62141317
                     rho |  .83930267   (fraction of variance due to u_i)
------------------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist3 lrgdpe_growth incumbent_lrgp
> de_growth, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       382
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1637                         Obs per group: min =        35
       between = 0.0300                                        avg =      54.6
       overall = 0.0542                                        max =        90

                                                Wald chi2(5)       =     86.06
corr(u_i, Xb)  = 0.0459                         Prob > chi2        =    0.0000

                                       (Replications based on 69 clusters in election_id)
-----------------------------------------------------------------------------------------
                        |   Observed   Bootstrap                         Normal-based
     positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
------------------------+----------------------------------------------------------------
              incumbent |    .182837   .0997785     1.83   0.067    -.0127252    .3783992
           incumbent_pm |   .2275542   .0777686     2.93   0.003     .0751306    .3799779
             extremist3 |  -.1625426   .0309142    -5.26   0.000    -.2231333   -.1019519
          lrgdpe_growth |  -.0017827   .0117096    -0.15   0.879    -.0247332    .0211678
incumbent_lrgpde_growth |   .0068686    .020363     0.34   0.736    -.0330421    .0467793
                  _cons |   2.008918   .1908028    10.53   0.000     1.634951    2.382885
------------------------+----------------------------------------------------------------
                sigma_u |  1.3984793
                sigma_e |  .63562656
                    rho |   .8287876   (fraction of variance due to u_i)
-----------------------------------------------------------------------------------------

. bootstrap, reps(400) seed(10101) cluster(election_id): xtreg positive_sentiment incumbent incumbent_pm extremist3 inflation_l incumbent_inflat
> ion_l unemployment_l incumbent_unemployment_l lrgdpe_growth incumbent_lrgpde_growth, fe;
(running xtreg on estimation sample)

Bootstrap replications (400)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
..................................................    50
..................................................   100
..................................................   150
..................................................   200
..................................................   250
..................................................   300
..................................................   350
..................................................   400

Fixed-effects (within) regression               Number of obs      =       359
Group variable: language                        Number of groups   =         7

R-sq:  within  = 0.1807                         Obs per group: min =        30
       between = 0.1534                                        avg =      51.3
       overall = 0.0074                                        max =        90

                                                Wald chi2(9)       =     90.87
corr(u_i, Xb)  = -0.1229                        Prob > chi2        =    0.0000

                                        (Replications based on 62 clusters in election_id)
------------------------------------------------------------------------------------------
                         |   Observed   Bootstrap                         Normal-based
      positive_sentiment |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
-------------------------+----------------------------------------------------------------
               incumbent |   .0575372   .1480368     0.39   0.698    -.2326095     .347684
            incumbent_pm |   .2144051   .0759793     2.82   0.005     .0654883    .3633219
              extremist3 |    -.15531   .0305586    -5.08   0.000    -.2152038   -.0954162
             inflation_l |  -.0276756   .0126919    -2.18   0.029    -.0525513      -.0028
   incumbent_inflation_l |   .0254406   .0155874     1.63   0.103    -.0051102    .0559913
          unemployment_l |  -.0222043   .0131991    -1.68   0.093    -.0480741    .0036654
incumbent_unemployment_l |  -.0043768   .0151246    -0.29   0.772    -.0340206     .025267
           lrgdpe_growth |  -.0131118   .0125594    -1.04   0.296    -.0377278    .0115043
 incumbent_lrgpde_growth |   .0103965   .0188697     0.55   0.582    -.0265874    .0473805
                   _cons |   2.334673   .2433138     9.60   0.000     1.857787    2.811559
-------------------------+----------------------------------------------------------------
                 sigma_u |  1.4564207
                 sigma_e |  .62043444
                     rho |  .84639924   (fraction of variance due to u_i)
------------------------------------------------------------------------------------------

. *     ****************************************************************  *;
. *       Replication complete                                            *;
. *     ****************************************************************  *;
. log close;
      name:  <unnamed>
       log:  C:\Users\mgolder\Dropbox\campaign_sentiment_paper\replication\data_analysis\sentiment2.log
  log type:  text
 closed on:  14 Jan 2019, 15:30:21
------------------------------------------------------------------------------------------------------------------------------------------------
